fusetools.stat_tools.Test

class fusetools.stat_tools.Test

Bases: object

Functions for implementing Statistical tests.

Methods

chi_squared_result	Calculates correlation between 2 proportions using a Chi-Squared test..
correlation	Performs a Pearson test of correlation between two data samples.
cramers_corrected_stat	Calculates correlation between 2 categorical variables using Cramer’s method.
pe	Calculates the Price Elasticity of Demand.
poisson	Performs a Poisson test which tests statistical difference between groups comparing counts over a period of time.
sample_size1	Calculates sample size needed for desired measuring effect size.
survival_result	Performs a survival test which tells if statistical difference in times until an outcome between two samples.
ttest	Performs a t-test between two groups split by a flag.
ttest_result	Performs a T-Test between two groups of data.

classmethod chi_squared_result(sample1_successes, sample1_trials, sample2_successes, sample2_trials)

Calculates correlation between 2 proportions using a Chi-Squared test..

Parameters

• sample1_successes – Sample 1’s successes.

• sample1_trials – Sample 1’s trials.

• sample2_successes – Sample 2’s successes.

• sample2_trials – Sample 2’s successes.

Returns

Chi-Squared p-value.

classmethod correlation(sample1_dat, sample2_dat)

Performs a Pearson test of correlation between two data samples.

Parameters

• sample1_dat – Sample 1 data array/list.

• sample2_dat – Sample 2 data array/list.

Returns

Pearson correlation result.

classmethod cramers_corrected_stat(cat_col1, cat_col2)

Calculates correlation between 2 categorical variables using Cramer’s method.

Parameters

• cat_col1 – Categorical column 1.

• cat_col2 – Categorical column 2.

Returns

Correlation between 2 categorical variables using Cramer’s method.

classmethod pe(type, original_quantity=False, new_quantity=False, original_price=False, new_price=False, pe_prices=False, pe_quantities=False)

Calculates the Price Elasticity of Demand.

Parameters

• type – Classification of whether data is in array/list data format or a scalar format (sample or other).

• original_quantity – Starting quantity demanded if data is scalar values.

• new_quantity – Ending quantity demanded if data is scalar values.

• original_price – Starting price if data is scalar values.

• new_price – Ending price if data is scalar values.

• pe_prices – Array/list of prices paid for quantities demanded.

• pe_quantities – Array/lust of quantities demanded.

Returns

Price elasticity of demand (float).

classmethod poisson(sample1_events, sample1_days, sample2_events, sample2_days)

Performs a Poisson test which tests statistical difference between groups comparing counts over a period of time.

Parameters

• sample1_events – Count of sample 1 events.

• sample1_days – Count of sample 1 days.

• sample2_events – Count of sample 2 events.

• sample2_days – Count of sample 2 days.

Returns

P-value for a Poisson statistical test.

classmethod sample_size1(baseline_input, effect_size_input, significance_level_input, statistical_power_input)

Calculates sample size needed for desired measuring effect size.

Parameters

• baseline_input – Baseline rate to measure effect against against.

• effect_size_input – Desired effect size to measure.

• significance_level_input – Desired level of statistical significance.

• statistical_power_input – Desired level of statistical power.

Returns

Calculated sample size.

classmethod survival_result(data_type, sample1_dat_survival=False, sample2_dat_survival=False, survival_confidence_level=False, sample1_dat_survival_mean=False, sample1_dat_survival_size=False, sample2_dat_survival_mean=False, sample2_dat_survival_size=False)

Performs a survival test which tells if statistical difference in times until an outcome between two samples.

Parameters

• data_type – Classification of whether data is in array/list data format or a scalar format (sample or other).

• sample1_dat_survival – Sample 1 data if array/list.

• sample2_dat_survival – Sample 1 data if array/list.

• survival_confidence_level – Confidence interval to assess measure test.

• sample1_dat_survival_mean – Sample 1 mean if scalar value.

• sample1_dat_survival_size – Sample 1 size if scalar value.

• sample2_dat_survival_mean – Sample 2 mean if scalar value.

• sample2_dat_survival_size – Sample 2 size if scalar value.

Returns

P-value for statistical significance in difference in times until an outcomes between two samples.

classmethod ttest(df, grp_col, grp_1_flag, grp_2_flag, target_kpi)

Performs a t-test between two groups split by a flag.

Parameters

• df – Pandas DataFrame containing data.

• grp_col – Column used to group the data.

• grp_1_flag – Value used to distinguish group 1.

• grp_2_flag – Value used to distinguish group 2.

• target_kpi – Column for the target metric to compare test across groups.

Returns

T-Test p-value.

classmethod ttest_result(sample1_dat_ttest, sample2_dat_ttest)

Performs a T-Test between two groups of data.

Parameters

• sample1_dat_ttest – Sample 1 dataset.

• sample2_dat_ttest – Sample 2 dataset.

Returns

T-Test p-value.